Editor | On 05, Jul 2000

Darrell Mann
Industrial Fellow
Department of Mechanical Engineering
University of Bath
Bath, BA2 7AY, UK
Phone: +44 (1225) 826465
Fax: +44 (1225) 826928
E-mail: D.L.Mann@bath.ac.uk

Introduction

This article is about how TRIZ can be used to design better wrenches, it is about defining the right problem to solve, and it is about using the difference between conceptual and specific solutions to transfer ‘good’ principles from one discipline to another.

The Wrench

Many of us have struggled to undo an over-tightened or corroded nut at some point in our lives. Conventional wrenches are not well-suited to such situations, and chances are that they will damage the nut in some way, and make it even more difficult to remove. This article was precipitated by a demonstration of a novel wrench which was able to undo a conventionally tightened, standard hexagonal nut which had had its corners systematically removed by an angle-grinder.

That wrench is of the close-end variety. The article will examine the design in more detail after first examining a different, open-ended wrench design which also seeks to eliminate the nut-damaging traits of conventional wrenches.

Open-Ended Wrench

Conventional wrenches damage nuts because the majority of the fastening or un-fastening loads are focused on the corners of the nut – Figure 1.

Figure 1: Conventional Wrench Focuses Loads on Nut Corners

If we look to TRIZ to improve the design of this simple, in evolutionary S-curve terms, relatively mature system, the Technical Contradiction problem solving tools offer a good start point. As we know from TRIZ findings, if we are to improve the design of a system at or close to the top of its current S-curve, we are going to have to identify and eliminate a contradiction. The damage causing traits of the current wrench are a symptom of the current (fundamental) limitations of the current design paradigm. These traits give us a good lead in helping to identify what contradictions we should tackle in order to improve the design. If we are to tackle this contradiction successfully, we need then to map our ‘reduce the likelihood of damage to the nut’ desire onto one or more of the parameters in the Contradiction Matrix (1). ‘Object generated harmful factors’ – parameter 31 – thus appears as a very good correlation. So, we know what it is we want to improve, we now need to work out what gets worse as we reduce the nut damage. This is often a more difficult task than identifying the parameter we are trying to improve. One common strategy is to work through the list of 39 parameters until one appears to fit the problem. Another, more systematic, approach involves a process by which we force ourselves to put all the solution constraints we know to exist to one side. A good question to ask is ‘how might I achieve the improvement goal if there were no obstacles?’

Asking such a question for the Figure 1 wrench, we might come up with answers like:-

a)I would make the sides of the wrench fit the sides of the nut exactly so that all of the flat face of the wrench touches all of the flat face of the nut.

b)I would add a ‘gizmo’ that allowed me to move the sides of the wrench so that they fitted the sides of the nut (or a gizmo that allowed the sides of the wrench to move by themselves)

c)I would make the wrench out of a softer material so that it was damaged instead of the nut

None of these is anything that ‘gets worse’ per se about the design of the wrench, so in order to be able to use the Contradiction Matrix, we need to map the answers to a generic worsening parameter from the list of 39. The first of the three answers – make the sides of the wrench fit exactly – is probably the most pragmatically practical one. In terms of relating the idea to ‘something that gets worse’, we might well here make a correlation between our desire for very good dimensional accuracy to the ‘manufacturability’ parameter (i.e. if I make the manufacture tolerance tighter, the manufacturability becomes worse).

We now have a very good technical contradiction:-
Thing I am trying to improve:	OBJECT GENERATED HARMFUL FACTORS
Thing which gets worse:	MANUFACTURABILITY
Matrix recommends:	4 ASYMMETRY
	17 ANOTHER DIMENSION
	34 DISCARDING AND RECOVERING
	26 COPYING

From Another Dimension we get ‘move an object in two or three-dimensional space’ or ‘use a different side of the given area’, from Discarding & Recovering, we get ‘make portions of an object that have fulfilled their function go away’, and from Asymmetry, ‘if an object is asymmetrical, increase the degree of asymmetry’. All of which give very strong pointers to the solution found in US Patent 5,406,868 from 1995 (Figure 2).

Figure 2: US Patent 5,406,868 ‘Open End Wrench’

(NB: This invention represents a step beyond the conceptually similar design currently marketed by ‘Metrinch’)

The motivation for this (and similar) inventions is precisely to eliminate the nut-damaging harmful side-effects of all conventional wrenches. The 5,406,868 solution achieves this objective by profiling the working faces of the wrench such that the points of contact with the nut avoid the damage-prone corners.

Trends of Evolution

The above description hopefully shows how the TRIZ Contradictions tools can be used to derive good conceptual solutions. A look at the TRIZ Trend relating to evolution of geometric structures – Figure 3 – gives us an alternative method of getting to the same end point.

These trend patterns are arranged such that ‘benefits’ increase as we pass from left to right along the trends. While it may not be immediately obvious that a ‘benefit’ occurs at each step – for example, without having seen the 5,406,868 solution and the preceding contradictions discussion, we might not have seen any benefit in moving from a ‘straight line’ to a non-straight line on the nut contact faces of the wrench – the trends are there to tell us that there is a benefit to be gained if we actively look for it.

Figure 3: ‘Geometric Evolution of Linear Structures’ Evolution Trend

(picture from TOPE^®)

Closed-End Wrench

Moving from an open-ended to a closed ended wrench design – especially if we include the good design practices identified in the Figure 2 or Metrinch designs – largely overcomes the objective generated harmful factors found in the open-ended design. We might consider this closed-end wrench design to be a good solution. TRIZ however is there to help prompt us to look for ever better solutions, and in particular to continue to tackle the contradictions present in any design – see ‘Contradiction Chains’ (Reference 1). Examining the Figure 4 design below, we might observe, for example, the fact that although theoretically the loads in the wrench and on the nut being tightened or loosened are distributed evenly on each contact face of the nut, in practice, irregularities in manufacture mean that the loads are not evenly spread. As with the previous discussion of open-ended wrench designs, this uneven load profile can cause damage to either the wrench or – more likely – the nut.

Figure 4: Closed-End Wrench Design, US Patent 4,930,378

In this case, we might see that in beginning to formulate a new contradiction, the parameter we are trying to improve is the stress distribution around the nut. In terms of the Contradiction Matrix, what we are trying to improve is TENSION, PRESSURE.

Thinking next about what gets worse as we try to improve the stress distribution, we see immediate parallels with the previous open-ended wrench discussion in that, we see that tightening the manufacture tolerances will allow us to assist the stress distribution. Thus we have a TENSION, PRESSURE versus MANUFACTURABILITY contradiction. Looking then to the Matrix we have:-

Thing I am trying to improve:	TENSION, PRESSURE
Thing which gets worse:	MANUFACTURABILITY
Matrix recommends:	1 SEGMENTATION
	35 PARAMETER CHANGES
	16 PARTIAL OR EXCESSIVE ACTION

Both Inventive Principle Number 1, Segmentation and Number 35 Parameter Changes (‘increase the degree of flexibility’) have been combined in the novel solution illustrated in Figure 5.

Figure 5: Liquid Levers Advanced Wrench Design

The novel design overcomes the contradiction by allowing the wrench to flex in such a way that the loads and stresses around the nut tend to become equalised. Thus a high load on one face causes a high degree of flexure and consequent re-distribution of loads to other faces.

The recommendation of the ‘Segmentation’ solution trigger is particularly interesting in the light of the otherwise non-instinctive step of introducing a cut into the closed wrench: ‘conventional’ design thinking would tend to suggest that segmenting a fundamentally strong closed structure would inevitably weaken it. In practice, of course, the novel wrench design overcomes the decrease in structure strength by adding an existing resource, such that now the nut becomes a part of the structural system. (Note: the invention is also a very good example of the identification and use of previously un-exploited resources in a system.)

The recommendation of Parameter Changes, and particularly the accompanying trigger ‘increase the degree of flexibility’, is likewise non-instinctive in this case. And yet the solution offered by adopting the Principle is considerable in terms of overcoming the stress contradiction.

Trends of Evolution Again

Looking at the Figure 5 wrench design might also cause us to reflect on the TRIZ dynamisation trend (Figure 6). The wrench profile may be seen to have evolved from an ‘immobile’ system to a ‘many jointed’ structure (in that the majority of the flexure built into the design occurs at the reduced cross-section areas around the wrench head).

The trend also suggests further wrench evolution opportunities.

Figure 6: ‘Dynamisation’ Evolution Trend

Conceptual Solutions

The way in which our minds work makes it far more likely that we will remember specific solutions before generic functions. This is a continual ‘psychological inertia’ problem.

A sometimes helpful strategy in this scenario is to extract from those ‘good solutions’ we see, good generically applicable conceptual strategies.

In the case of the earlier open-ended wrench solution, the good ‘generic concept’ is keep loads away from areas which are not well suited to carrying them.

In the case of the subsequent closed wrench solution, the good ‘generic concept’ is all materials have some inherent deflection when placed under load; use these deflections to perform a useful function. (See also Inventive Principle 22 ‘Blessing In Disguise’.)

Similarly, the flange joint article from some time ago (Reference 2), also allows us to extract good generic conceptual solutions from one specific solution; when trying to perform a sealing function, try to do this over a small contact area. (See also the Space Shuttle Challenger investigation for another example.)

Conclusions

1)The TRIZ Contradictions tools are powerful design paradigm changers. The wrench example provides further evidence that, for simple mechanical systems at least, the Contradiction Matrix is highly likely to give very good solution pointers.

2)The Contradiction and Evolution Trends tools can often get us to the same solutions, albeit by different routes.

3)If we can extract generic concepts from good specific solutions, we have the basis for a powerful database of ‘good design practice’.

References

1)Mann, D.L., ‘Contradiction Chains’, TRIZ Journal, January 2000.

2)Mann, D.L., ‘Case Studies In TRIZ: Halving The Number Of Bolts Around A Flange Joint’, TRIZ Journal, October 1998.

The spanner concept described in this article was conceived by Nigel Buchanan of Liquid Levers Ltd in Scotland. More details about the spanner and Nigel’s other work may be found at the Liquid Levers web-site, www.liquidlevers.com.

ã2000, D.L.Mann, all rights reserved.